Applying cognitive interviewing process to refine an instrument for measuring deprescribing experience and outcomes in older veterans

Background Currently, there is no validated instrument that assesses a patient’s reported deprescribing experiences and outcomes. This study reports the third phase of refining such an instrument, the Deprescribing Evaluation and Quality Improvement instrument (DEQI), using cognitive interviews. Cognitive interviews have been used in the development of many survey tools; however, there is limited evidence of their use in pharmacy research, particularly in deprescribing studies involving older adults. The aim of this phase of the DEQI instrument development is to evaluate and refine the individual items to enhance their clarity to accurately reflect patient-reported experiences and outcomes after deprescribing interventions. Methods A draft DEQI Instrument was tested through two rounds of cognitive interviews conducted four weeks apart, involving 16 Veteran adults aged 65 or older who were on multiple medications. Interviews lasted about 60–90 min. Two cognitive interviewing techniques, “think-aloud” and “verbal probing,” were used. Data from the interviews were independently analyzed by four coders employing two coding techniques: interaction coding and the cognitive classification scheme (CCS) to identify problematic questions in the DEQI instrument. Based on the analysis of the first round, the items were revised and then administered again in a second round. The improvement in clarity was evaluated. Results Participants were mostly male (88%) and non-Hispanic White (69%), with a mean age of 73.4 years (range 64–84). About 38% had some college or technical school education. The classification coding scheme identified 102 problematic items in the first round, with approximately 70% related to the comprehension and communication code, and around 14% related to response selection. After modifications and readministration in the second round, the number of identified problems dropped significantly from 102 to 21. Following the revision in the second round, the proportion of “immediately codable” answers increased significantly (from 0.68 to 0.80), while the proportions of “uncodable” answers and “items requiring clarification” decreased. Conclusions This study demonstrates the use of interaction coding and CCS to refine an effective evaluation instrument to improve deprescribing interventions with an older Veteran population. Through iterative refinement of items based on cognitive interviews, over half of the DEQI instrument’s items were improved, resulting in clearer items and reduced confusion among participants. Clinical trial number Not applicable.